After a dropped call, three stories appear at once. The internet provider is congested. The communication app “has been bad this week.” The cheap headset is rubbish. One of those is sometimes true. All three are rarely true on the same afternoon.
Handset first, because it is the cheapest test. Swap the headset. Try the same app on a second phone sitting still on the desk. If the drop follows the first handset, stop. Charge the battery, check whether the phone is also downloading an update, and only then talk about the building.
App next. Place the same conversation on a different app for one day — WhatsApp instead of the in-house dialler, or a cellular call instead of Teams. If only one app dies, the path to that app’s servers, or the way that app hides loss, is the lead. We still measure; we do not guess from a forum post about an outage in another country.
ISP last, because it is the conversation nobody wants to have without paper. If every app dies at the same minute, and a phone on mobile data in the stairwell is clean, the building’s outgoing link is the story. A packet loss assessment gives you timestamps. Timestamps survive a support chat that resets every Friday.
Shared offices in the Klang Valley add a fourth suspect: the floor’s common Wi-Fi, named after the building, used by three companies at lunch. Your lease may not mention voice. Measure anyway. The finding still helps you ask for a quieter hour, a different access point, or permission to use a dedicated socket.
We will not pick a villain to make the readout satisfying. We will say which test failed, at which minute, on which path. That is slower than a slogan and much harder to argue with.